Respiratory Microbiome Profiles Associated With Distinct Inflammatory Phenotype and Clinical Indexes in Chronic Obstructive Pulmonary Disease.

Introduction/objective: Respiratory microbiome studies have fostered our understanding of the various phenotypes and endotypes of heterogeneous chronic obstructive pulmonary disease (COPD). This study aimed to identify microbiome-driven clusters that reflect the clinical features and dominant microbiota of COPD.

Methods: This cross-sectional study included 32 patients with stable COPD between December 2019 and December 2020 from the outpatient clinic of the China-Japan Friendship Hospital. Sputum samples were tested for 16S rRNA. Patients were classified according to the species level using an unsupervised clustering method to compare the inflammatory phenotypes of 2 clusters and analyze the correlation between the main bacteria and clinical indicators in each cluster. Patients were further divided into 2 clusters according to microorganisms.

Results: Neutrophils in cluster 1 were significantly increased compared with cluster 2. Cluster 1 was predominantly Bacteroides, while cluster 2 was dominated by Prevotella and Fusobacterium at the genus level. Fusobacterium was negatively correlated with the COPD Assessment Test (CAT) score, and Bacteroides were positively correlated with the number of acute exacerbations of COPD.

Conclusion: This study found that differential flora was negatively associated with CAT scores and the number of acute exacerbations of COPD. This microbiome-driven, unbiased clustering method for COPD can help identify new endotype-related COPD phenotypes.